Despite these achievements in iNSC technology little is know
Despite these achievements in iNSC technology, little is known on how cellular identities of iNSCs differ from those of NSCs in comparison to pluripotent stem cell-derived NPCs. For instance, it is currently unknown to what extent neural, metabolic and stress response pathways differ between these neural cell populations, although ESC–NPCs have previously been included as a neural cell population in a comparative study on iNSCs (Cassady et al., 2014). It is also not known if positional neural cell identities in iNSCs are irreversibly established or if they can be shifted along the rostro-caudal or ventro-dorsal axis as demonstrated for ESC-derived NPCs (Bertacchi et al., 2013). Furthermore, data are lacking on whether neural positional marks in iNSCs remain stable or change after transplantation into the adult rodent brain. These data could have significant implications for biomedical applications using iNSCs such as disease modeling or cell replacement therapy since cell metabolism definition with appropriate neural programs may have to be derived for these purposes. To address these questions, we have derived iNSCs from fibroblasts by transcription factor-mediated transdifferentiation and performed comparative studies on iNSCs both in vitro and in vivo.
Materials and methods Please see the Supplemental Information for further details. Data of at least three independent experiments are presented as mean+SEM. Statistical significance was determined by unpaired Student\'s t-test.
Discussion While iNSCs, NSCs and ESC–NPCs share global expression marks distinct from those in fibroblasts (Cassady et al., 2014), data is lacking on how cellular programs in these neural progenitor populations relate to each other. Our comparative transcriptome analysis revealed distinct global and neural profiles in iNSCs, which had higher similarities to those in brain-derived NSCs than to those in ESC–NPCs. Similar programs were also seen when comparing metabolic, stress- and cell cycle-associated marks in iNSCs, NSCs and ESC–NPCs. These findings could have significant implications for the use of iNSCs in biomedical research, as the close resemblance of self-renewing iNSCs and NSCs could be exploited for drug and toxicology screening assays or for disease modeling purposes. In fact, iNSCs may constitute a suitable, time- and cost-effective alternative cell source to pluripotent stem cell-derived neural cells when attempting characterization or modification of disease-related processes in brain-NSC-like cells and their progeny as, for instance, in neurodegenerative diseases. The potential of this approach is further heightened by the fact that iNSCs with neuronal and glial differentiation potential have recently been generated also from human fibroblasts using SOX2 alone, though challenging, or using SOX2 in combination with HMGA2/anti-let7b (Ring et al., 2012; Yu et al., 2015). Our data thus strongly encourage similar, thorough comparisons of neural cell identities in human NSCs, human iNSCs and in human pluripotent stem cell-derived NPCs, especially of patient-derived iNSCs and iPSC–NPCs and their derivatives, to identify optimal human stem cell populations for abovementioned screening and disease modeling purposes. By following a similar approach, we recently compared human iPSC–NPCs from different somatic origins and could demonstrate distinct origin-dependent neural programs in these cells (Hargus et al., 2014a). An analysis of positional marks along the rostro-caudal axis revealed a hindbrain/posterior cell identity in iNSCs, which could be further shifted towards caudal and partially towards rostral but not towards ventral fates in vitro. Hence, our iNSCs may not be well suited to generate floor plate cells in vitro as similarly seen for our forebrain-derived NSCs that were cultured in parallel. On the other hand, iNSCs readily acquired caudal marks and might thus be applicable for repair of caudal neural domains such as the spinal cord. In line with this observation, transplanted iNSCs promoted functional recovery in a rodent model of spinal cord injury when generated by an identical protocol (Hong et al., 2014) or by a protocol using Sox2 alone for direct neural conversion (Liu et al., 2015).